Method development and application to determine potential plant uptake of 1

antibiotics and other drugs in irrigated crop production systems 2 3

4

Tammy L Jones-Lepp*1

, Charles A Sanchez2, Thomas Moy

3, Reza Kazemi

4 5

1U.S. Environmental Protection Agency, Research Chemist, Office of Research and 6

Development, National Exposure Research Laboratory-Environmental Sciences Division, 7

Las Vegas, NV 89119, (702) 798-2144, jones-lepp.tammy@epa.gov; 2University of 8

Arizona, Department of Soil, Water, and Environmental Sciences, Yuma Agricultural 9

Center, Yuma, AZ; 3Senior Environmental Employee Program, U.S. Environmental 10

Protection Agency, 944 E. Harmon Ave., Las Vegas, NV 89119; 4Student Services 11

Contract, U.S. Environmental Protection Agency, 944 E. Harmon Ave., Las Vegas, NV 12

89119 13

14

15

16

17

18

Abstract 19 20

Studies have shown the detection of emerging contaminants (ECs), of which 21

pharmaceuticals are a subset, in surface waters across the United States. The objective of 22

this study was to develop methods, and apply them, to evaluate the potential for food 23

chain transfer when EC-containing waters are used for crop irrigation. Greenhouse 24

experiments were performed where select food crops were irrigated with water spiked 25

with three antibiotics. Field experiments, at two different sites, were conducted. Select 26

crops were irrigated with wastewater effluent known to contain ECs, EC-free well water, 27

and Colorado River water containing trace-level ECs. The results of the greenhouse 28

studies show the potential for uptake of one or more of the antibiotics evaluated, albeit at 29

very low levels. In those food crops watered with wastewater effluent only an industrial 30

flavoring agent, n,n’-dimethylphenethylamine (DMPEA) was consistently found. None 31

of the evaluated contaminants were found in crops irrigated with Colorado River water. 32

33

34

35

Keywords: emerging contaminants; crop uptake; pressurized liquid extraction; liquid 36

chromatography-electrospray-ion trap mass spectrometry/mass spectrometry; LC-MS/MS 37

38

Introduction 39

In the southwestern part of the United States, increasing demands on scarce water 40

resources has forced water authorities to look for alternative water resources. Some 41

water authorities use treated wastewater effluent for injection into ground water aquifers 42

for the purpose of pumping it out later and re-use, with further treatment, as drinking 43

water (1). Other municipalities use treated wastewater effluent for non-potable water 44

reuse, e.g., watering of golf courses and municipal green spaces, as well as a source of 45

irrigation water for crops (2). Of concern are the reports of numerous pharmaceuticals 46

and other emerging contaminants (ECs) found in these groundwaters. Rowe et al. (3) 47

reported that at least one EC was present in 76% of shallow urban wells sampled in the 48

Great and Little Miami River Basins in Ohio and found that the number of ECs detected 49

increased with increasing urban land use. 50

51

Although pharmaceuticals designed for human or veterinary use have a specific 52

biological mode of action, the impact on non-target species is rarely known. Since 53

pharmaceuticals are released into the environment as complex mixtures, and not as 54

individual compounds, there exists the possibility for synergistic, or antagonistic, 55

interactions resulting in unexpected biological effects. The concentrations of 56

pharmaceuticals in drinking water supplies are likely to be below any level of direct risk 57

to humans. However, it is the persistence and presence of antibiotics in the environment 58

that could pose a serious threat to human health. (4-7). The principal existing concern 59

with antibiotics is the identification of growing resistance in microbial populations (7-60

10). Resistance has been found in bacteria isolated from the innards of animals treated 61

with antibiotics, in their corresponding manure (11), and in agricultural soils receiving 62

manure (12, 13). There is concern that non-pathogenic bacteria can serve as a platform 63

for gene transfer to pathogenic organisms as a result of promiscuous exchange of genetic 64

material among microbes (5, 14). Antibiotic-resistant bacteria have been found in surface 65

water (6, 7), sediments (15, 16), and ground water (10, 17). 66

67

Recent studies have shown that human-use antibiotics (azithromycin, 68

clindamycin, and roxithromycin) are environmentally available in wastewaters, source 69

waters, and biosolids (18-21). Several researchers have demonstrated that certain 70

veterinary antibiotics (e.g., florfenicol, trimethoprim, sulfamethazine, enrofloxacin, etc.) 71

can be taken up into food crops (e.g., wheat, corn, lettuce, barley, and potato) produced 72

on manure-amended soils (22-25). Recently, Herklotz et al. (26) published a study of the 73

uptake of human pharmaceuticals (e.g. carbamazepine, salbutamol, sulfamethoxazole, 74

and trimethoprim ) into cabbage (Brassica rapa var. pekinensis) and Wisconsin Fast 75

plants (Brassica rapa) in a hydroponic garden setting. 76

77

At the Imperial Diversion Dam (IDD) near Yuma, AZ almost 5 billion m3 of 78

water are diverted from the Colorado River to irrigate the approximately 400,000 ha of 79

agricultural crops that are shipped nationally and internationally. Previous research has 80

shown that the Colorado River is contaminated with low levels of perchlorate and this 81

contaminant can be detected in most agricultural commodities irrigated with this water 82

(27, 28). Macrolide antibiotics, pseudoephedrine, and illicit drugs have been identified in 83

several municipal wastewater streams that discharge into the Colorado River (29). There 84

is a probability that the drugs present in water could potentially reach food crops. The 85

research presented in this paper will focus on the development and ground-truthing of 86

analytical methods for determining the fate of ECs (e.g., antibiotics, illicit drugs, over-87

the-counter (OTC) drugs) into food crops via a three-part study. There was an emphasis 88

on method development for detecting three antibiotics - azithromycin, roxithromycin and 89

clindamycin. Azithromycin and clindamycin due to their wide-spread usage in the US 90

(18), 91

http://drugtopics.modernmedicine.com/drugtopics/data/articlestandard//drugtopics/2520192

0/674976/article.pdf, and roxithromycin due to its surreptitious usage. Roxithromycin is 93

not prescribed in the US, but has been detected in wastewaters and biosolids in the US 94

(18, 20). 95

96

MATERIALS AND METHODS 97 98

Chemicals. Clarithromycin was obtained from U.S. Pharmacopeia (Rockville, 99

MD). Azithromycin, roxithromycin, clindamycin, and n,n’-dimethylphenethylamine 100

(DMPEA), were obtained from Sigma-Aldrich (St. Louis, MO). Methamphetamine, 101

MDMA, d5-MDMA, and pseudoephedrine were obtained from Cerilliant Corporation 102

(formerly Radian Corp., Round Rock, TX). HPLC-grade methanol was obtained from 103

varying sources [e.g., Burdick and Jackson (Muskegon, MI); EK Industries (Joliet, IL); 104

JT Baker (Phillipsburg, NJ)]. Acetic acid, glacial ACS reagent grade (VWR, West 105

Chester, PA); acetonitrile (Burdick and Jackson, Muskegon, MI); formic acid ACS 106

reagent grade (Anachemia, Rouses Point, NY); methyl tertbutyl ether (MTBE) (VWR, 107

West Chester, PA); and deionized water (NANOpureTM

, Barnstead, Dubuque, IA). 108

109

Stock standard solutions were individually prepared in HPLC-grade methanol and 110

stored in the dark at 4oC. A high-level standard mix (containing the macrolide antibiotics 111

and the other drugs/chemicals), at concentrations of 10 or 20 ng µL-1

, was prepared 112

monthly in methanol, and a calibration standard mix was prepared weekly at 113

environmentally relevant concentrations (0.5 to 1 ng µL-1

) in 99% methanol:1% acetic 114

acid. 115

116

Samples 117

Multiple samples, e.g., soils, waters, plants, were collected and processed during the three 118

phases of the study. A brief summary of the samples collected and their sources are 119

listed in Table 1. 120

121

Phase I - Greenhouse study, plant materials and growth conditions. The first 122

phase of the study was a controlled greenhouse experiment. Three crops, lettuce 123

(Lactuca sativa), spinach (Spinacia oleracea), and carrots (Daucus carota sativus), were 124

initially germinated in potting soil and irrigated with unspiked Colorado River water. At 125

approximately the four-leaf stage, the plants were transplanted into 1.5 L pots filled with 126

1.5 kg of washed silica sand. From transplanting through harvest, the plants were 127

irrigated with Colorado River water spiked with varying concentrations of three 128

antibiotics: azithromycin, clindamycin, and roxithromycin. The antibiotics were 129

dissolved in a small amount of methanol, and then diluted to 1000 ng L-1

with Colorado 130

River water. All concentrations were achieved by serial dilutions with Colorado River 131

water. The dosing concentrations were selected relative to concentrations found in 132

wastewater effluent streams (29) and were dosed at 0 (control), 0.1, 1, 10, 100, and 1000 133

ng L-1

. It was observed that the concentrations of the macrolide antibiotics in the 134

prepared irrigation water declined with time, perhaps due to photodegradation, microbial 135

degradation, or adhesion to the walls of the plastic container. Thus, solutions were 136

prepared weekly to maintain the target concentrations. Four replicates, at each 137

concentration, were performed. Plant selection per treatment was done following a 138

complete randomized design. After harvest, the crop plants were partitioned into leaves 139

and roots and then frozen. The frozen samples were freeze-dried, and weights before and 140

after freeze-drying were recorded. The freeze-dried samples were ground and stored in 141

vials for later extraction. 142

143

Phase II - Field studies UA-CAC. The second phase of the study was to ground-144

truth the methods developed during the first phase. This phase of the study was 145

conducted at the University of Arizona Campus Agricultural Center (UA-CAC), Tucson, 146

AZ. This was accomplished by applying the developed methods to field-grown crops 147

irrigated with treated City of Tucson wastewater effluent that contained known amounts 148

of ECs and, as a control, irrigated with well water known to be EC-free (Table 2). The 149

growing field consisted of loam-textured soils and was split into two separate sections. 150

The first half was irrigated by furrows filled with treated wastewater effluent and the 151

other section, the control, was irrigated by furrows filled with well water (Table 3). On 152

March 10, 2008, peppers (Capsicum annuum), tomatoes (Lycopersicon esculentum), 153

melons (Cucumis melo), lettuce, and watermelon (Citrullus lanatus) transplants were 154

planted in raised beds on 1-m centers. Spinach and carrots were seeded in these same 155

beds. Identical crops were established in each of the two sections. The crops were 156

fertilized and pests were controlled using standard practices. The crops were irrigated as 157

needed and harvested as each crop species matured. The final harvest was June 15, 2008. 158

After harvest, the crop plants were partitioned into leaf and root segments, and where 159

appropriate fruit, and frozen. The frozen samples were subsequently freeze-dried, and 160

weights before and after freeze-drying were recorded. The freeze-dried samples were 161

ground and stored in vials for later extraction. 162

163

Phase III - Field studies UA-YAC. During the third phase of the study, the 164

same crops as used in the Tucson studies were grown and collected at the University of 165

Arizona-Yuma Agricultural Center (UA-YAC), Yuma, AZ. All crops were grown on 166

loam-textured soils and irrigated with Colorado River water diverted at the IDD, north of 167

Yuma (Table 3). An opportunity arose to sample Bermuda hay grass from a field, close 168

to UA-YAC, that had a long-term history of application (several years’ worth) of EC-169

containing biosolids. The biosolids used on the field were obtained from Hyperion 170

wastewater treatment plant (WWTP), Orange County, CA, whose biosolids had 171

previously been characterized for ECs (20). The Bermuda grass samples were sampled 172

for the purpose of studying the possible migration of ECs from the biosolids into 173

Bermuda grass grown as feedstock for livestock. 174

175

Water samples. Phase I. Colorado River water, used in the greenhouse studies, 176

was sampled during each collection period. Phase II. Well water and treated wastewater 177

effluent used in the UA-CAC field study were sampled approximately every other 178

irrigation period. These water samples were kept on ice, or refrigerated, until processing. 179

Phase III. Water, which was diverted at the IDD for agricultural use in the Yuma region 180

of the lower Colorado River, was sampled monthly at the main Yuma conveyance siphon 181

during the crop-growing period of the field crops being sampled. 182

183

Water extractions. Water samples were prepared for analysis using solid phase 184

extraction (SPE) Oasis MCX cartridges (Waters Corp., Milford, MA) with an automated 185

extractor (AutoTrace, Caliper Life Sciences, Hopkinton, MA). Oasis MCX cartridges 186

were prepared for use by loading at a rate of 1mL min-1

, 5 mL each of methanol, 187

deionized water, and 95:5 water:methanol. All water samples were pH adjusted to < pH 188

3, with 12 N HCl, and 500 mL were passed through the prepared Oasis MCX cartridges 189

at a rate of 7 mL min-1

. The cartridges were then dried for 15 minutes (using N2), then 190

extracted with 5 mL of 90:10 MTBE/methanol, followed by 10 mL 191

methanol/4%ammonium hydroxide. The resultant extracts were reduced to 0.5 mL using 192

4 to 10 psi of nitrogen, via an automated evaporator (TurboVap, Caliper Life Sciences, 193

Hopkinton, MA). Sample extracts were analyzed by liquid chromatography-194

electrospray-ion trap mass spectrometry/mass spectrometry (LC-ESI-ITMS/MS). 195

196

Plant and soil extractions. Crop samples were freeze-dried for 48 hours, or 197

longer, until moisture was no longer present. The freeze-dried samples were ground to a 198

semi-fine state, such that they passed through a sieve size of 300 μm, and stored in vials 199

until extraction. 200

201

Test plot and field soil samples were poured into clean 2-liter beakers and air-202

dried. The dried soils were ground to ~ 300 μm using a high impact ball mill (Mixer Mill 203

301, Retsch Inc, Newtown, PA). 204

205

Pressurized liquid extraction (PLE) of plant and soil samples. One gram each 206

of prepared plant and soil samples was extracted using an Accelerated Solvent Extraction 207

(ASE) system (Model ASE 200 Accelerated Solvent Extractor, Dionex Corporation, 208

Sunnyvale, CA) in 22-mL stainless steel extraction cells according to the following 209

procedures. 210

211

Extraction cell preparation. A glass microfiber filter, 2 cm (Ahlstrom, Helsinki, 212

Finland) was placed at the bottom of the extraction cell. Dependent upon whether soils 213

or plants were to be extracted the extraction cell(s) were prepared as follows: 214

215

Soil sample extraction cell preparation. Three grams of fluorosil were added to the 216

cell, followed by a layer of 3 g of alumina. 217

218

Plant sample extraction cell preparation. Three grams of alumina were added to the 219

cell, followed by a layer of 3 g fluorosil. 220

221

The final sample cell preparation, whether soil and plant samples, was the same. 222

A mixture of 1 g of sample (soil or plant) and 1-g of HydromatrixTM

was added to the 223

extraction cell, followed by 3 g alumina. HydromatrixTM

was filled to top and the 224

extraction cell was capped with another glass microfiber filter and sealed. 225

226

PLE extraction procedure. A two solvent extraction regime was necessary in 227

order to fully extract the analytes from the solid matrices. The prepared cells were placed 228

into the ASE and initially extracted with a mixture of MTBE:methanol (90:10) and 229

flushed at 80% of cell volume. Temperature and pressure were kept steady at 50oC and 230

1500 psi, respectively. After a static period of 15 minutes, the eluant was purged into a 231

clean collection vial. The cells were left in situ, and further extracted with a mixture of 232

methanol/1%acetic acid and flushed at 80% of cell volume. The temperature and 233

pressure were maintained at 80oC and 2800 psi, respectively. After a static period of 15 234

minutes, the eluant was purged into a clean collection vial. 235

236

PLE extract concentration and cleanup. The MTBE/methanol extract was 237

placed into a TurbovapTM

tube and reduced to 5 mL, using 4 to 10 psi of nitrogen, via an 238

automated evaporator (TurboVapTM

, Caliper Life Sciences, Hopkinton, MA). The 239

methanol/acetic acid extract was then combined with the reduced MTBE/methanol 240

extract and evaporated until a combined extract sample volume of 5 mL was reached. 241

The 5-mL extracts were removed from the TurboVapTM

and washed with 1 to 2 mL of 242

hexane. The number of hexane washes varied from one sample to another, but typically 243

washes were done as many times as necessary to clean the sample of any undesirable 244

compounds, such as chlorophyllic compounds, fatty and waxy materials. The cleaned 245

extracts were placed back into the TurboVapTM

, further concentrated to 0.5 mL and 246

solvent exchanged with methanol/1% acetic acid before analysis by LC-ESI-ITMS/MS. 247

248

Validation of plant extraction method. The PLE method was validated by using 249

a modified extraction technique that had previously been published for extracting ECs 250

from biosolids (20). The spiked plant materials were extracted and analyzed by LC-ESI-251

ITMS/MS. The resultant accuracy and precision data are shown in Table 5. 252

253

LC-ESI-ITMS/MS analysis. 254

Liquid chromatography. Chromatographic separations were performed using an 255

Ascentis Express C18 (Supelco-Aldrich, Bellefonte, PA) 2.7 μm particle size, 3 cm x 2.1 256

mm column, coupled with a Varian guard column (MetaGuard 2.0 mm Pursuit XRs 3µm 257

C18). Compositions of the mobile phases were as follows: (A) deionized water/0.5% 258

formic acid and (B): 82% methanol/18% acetonitrile/0.5% formic acid. The flow rate 259

through the column was 200 μL min-1

, with the following gradient elution conditions: 260

mobile phase A 100%, hold for 2 min; 3 min gradient to 30% A:70% B, hold for 5 min; 3 261

min gradient to 100% A, hold for 2 min; end run, 5 min equilibration time between 262

analyses. 263

264

Mass spectrometry. Mass spectrometric data were acquired with an iontrap 265

mass spectrometer, Varian 500MS (Walnut Creek, CA USA), configured with a liquid 266

chromatograph and an electrospray ion source. The 500MS was run in the positive 267

ionization mode under the following conditions: 268

ES needle was 5 kV 269

Drying gas was set at 20 psi and 350°C 270

Housing chamber at 50oC 271

Nebulizer gas at 40 psi 272

Spray shield at 600V 273

Capillary voltages were set dependent upon the optimized response of the 274

product ions of interest. 275

276

The molecular weight of the ECs of interest, the precursor and product ions 277

formed under LC-ESI-ITMS conditions, and the mass spectrometric limits-of-278

detection of the ECs are listed in Table 4. Due to the large amounts of interfering 279

materials co-extracted with the ECs, the analyses were performed using the 280

collision induced dissociation (CID) mode for both identification and quantitation 281

of the analytes of interest (18). 282

283

RESULTS AND DISCUSSION 284

The steps in environmental method development involved: (1) the ability to 285

extract the analytes of interest with some degree of precision and accuracy from an 286

environmental matrix; and (2) the ability to accurately identify and measure at low 287

(environmentally relevant) concentrations the analytes of interest. The focus of the 288

results and discussion section is on the plant extraction procedures and the results of the 289

finalized plant extraction method as applied to the various plant samples. 290

291

Analytical challenges. During the development and execution of this 292

methodology for plants, various analytical difficulties were encountered, both in the 293

extraction phase and the detection phase. For example, chlorophyll, waxy and fatty 294

materials were co-extracted from plant materials, but they were not fully removed during 295

the hexane cleanup phase, even after multiple (4x) washes. Injection of plant and root 296

extracts into the mass spectrometer built up deposits on the inner spray shield, causing 297

loss of sensitivity and necessitating cleanup of spray shield after every second injection 298

of sample extracts into the mass spectrometer. 299

300

Injection of some plant and root extracts temporarily bound non-dissolvable 301

materials to the column, even with a guard column in place, resulting in poor 302

chromatography. This problem necessitated reversing the flow into the chromatographic 303

column. The column was flushed first with methanol/0.5% formic acid and then with 304

deionized water/0.5% formic acid before the column was usable again. 305

306

Results of water analysis. Phase I. All contaminants evaluated were below 307

detection in the Colorado River water collected for spiking in the greenhouse studies. 308

Phase II. The treated Tucson wastewater effluent, used at UA-CAC field studies, 309

contained the macrolide antibiotic azithromycin, the OTC drug pseudoephedrine, the 310

illicit drug methamphetamine, and an industrial compound, n,n’-DMPEA (an isomeric 311

compound to methamphetamine), Table 1. All contaminants evaluated were below 312

detection in the control well water used during the Phase II experiments at UA-CAC. 313

Phase III. Previous studies have found a number of ECs in wastewater discharged at 314

various points along the Colorado River (29). However, almost all ECs were below 315

levels of detection for Colorado River water that was collected at the IDD (main Yuma 316

irrigation siphon). The one exception was Ecstasy (MDMA), which was detectable but 317

not quantifiable during the warmer months (June through September). 318

319

Validation of PLE method. It is difficult to compare the recoveries of ECs from 320

crops in this study to the few other studies on plant uptake that have been published (22, 321

24, 26) because those studies did not indicate findings of percent recovery of spiked ECs. 322

Boxall (22) does briefly mention, “Although recoveries for most determinands were 323

good, low but reproducible recoveries were obtained for selected substances in soil 324

and/or plant material, so all measured values were recovery corrected. These low 325

recoveries were observed for the highly sorptive study substances.” The actual spiked 326

recovery data, however, was not published. Most recently, Herklotz et al. (26) reported 327

percent recoveries of spiked ECs from carrots and cabbages. Their method, similar to the 328

one reported in this paper, used PLE, and they reported > 70% recoveries of 6 different 329

ECs. However, their methodology used either a mass labeled internal standard 330

calibration or a combination of standard addition and mass labeled internal standard 331

calibration, to calibrate and calculate the percent recoveries. 332

333

In comparison to Herklotz et al.’s method(26) the method presented in this 334

research used external standard calibration with no corresponding mass labeled 335

compounds, for calibration and quantitation. The best recoveries of ECs (i.e., 336

azithromycin, roxithromycin, and clindamycin) from the plant materials were generated 337

by packing the extraction cell with a layer of alumina, followed by a layer of fluorosil. 338

With the PLE method reported in this article, the percent recoveries of the spiked ECs 339

were low, on average 25%-30% recovery, but reproducible, as measured by percent 340

relative standard deviation (RSD), most were < 17% RSD, Table 5. The EC amounts 341

detected in the non-spiked plant materials were spike-corrected using an equivalent 342

spiked matrix. 343

344

While the use of labeled compounds will give a sense of higher recoveries, in 345

truth, the labeled compound is correcting for the low recovery of the native compound. 346

One downside to the use of mass labeled compounds in these types of studies is the 347

usually higher costs (compared to non-labeled standards) associated with their purchase 348

and the lack of many of the ECs with an accordingly matched mass-labeled compound. 349

350

Results of plant uptake studies. Phase I Greenhouse - Above ground dry 351

matter production averaged 1.5, 3.3, and 1.9 g for the spinach, lettuce, and carrots, 352

respectively, and 2.0 g for carrot roots. There were no statistically significant differences 353

in dry matter production among the macrolide treatment rates indicating no phytotoxocity 354

to these macrolide antibiotics up to 1000 ng L-1

in irrigation water. The greenhouse study 355

indicated that there were traces of uptake of clindamycin into the spinach roots, lettuce 356

roots and carrot roots, Table 6. Trace amounts of roxithromycin were also detected in 357

lettuce roots and carrot roots. Carrots showed the greatest amount of uptake of 358

roxithromycin, an average of 110 ng g-1

, from the 1000 ng L-1

treatment. Neither 359

clindamycin nor roxithromycin were detected at the lower than 1000 ng L-1

treatments. 360

The greenhouse study demonstrates potential for EC uptake from contaminated irrigation 361

water. 362

363

Phase II – Field studies UA-CAC. The field study at UA-CAC was a side-by-364

side comparison and it did not include true replication so statistical evaluations of 365

production were not possible. However, the observed production was generally lower in 366

the plot receiving effluent compared to that receiving well water. Most of the crops 367

evaluated are sensitive to salinity; therefore, the high salinity (1.2 dS m-1

) in the effluent, 368

as compared to the well water (0.2 dS m-1

), may have caused the limited production. 369

370

Although several of the ECs studied were constantly present in the Tucson treated 371

wastewater effluent, Table 2, only n,n’-DMPEA was consistently found in the UA-CAC 372

food crops irrigated with wastewater effluent. No uptake of azithromycin was seen in 373

any of the plant/root samples from Tucson effluent field crops. No detectable levels of 374

the study pharmaceuticals were found in the soils collected from the root zones of the 375

crops sampled in the Tucson effluent field crops. 376

377

The results of the greenhouse study, and the field study with treated effluent 378

wastewaters, indicate a potential for uptake of pharmaceuticals from contaminated water, 379

albeit at very low levels. At present, it seems that the pharmaceuticals tested are 380

sufficiently diluted, or degraded, within the main channel of the Colorado River and that 381

risks of uptake by crops irrigated downstream of municipal waste discharges are minimal. 382

383

Phase III – UA-YAC and biosolids amended field. None of the ECs evaluated 384

were found in spinach crops grown in the UA-YAC fields irrigated with Colorado River 385

water. This was not surprising considering no detectable levels of these contaminants 386

were present in the Colorado River water diverted for irrigation at the IDD. However, in 387

previous studies, perchlorate accumulation has been found in plants where the 388

contaminant was not detectable in irrigation water (27), perhaps due to soil accumulation 389

or plant bioconcentration. No detectable levels of the study pharmaceuticals were found 390

in the soils collected from the root zones of the crops sampled in the Yuma area. 391

392

From a field nearby to UA-YAC, soils and Bermuda grass were collected. This 393

field had been treated for several years with biosolids from the Hyperion WWTP (Orange 394

County, California), and the Bermuda grass was being used for animal fodder. While 395

none of the ECs evaluated were detected in either the soils or Bermuda grass grown in 396

those soils, azithromycin, clarithromycin, and n,n’-DMPEA were detected in the roots of 397

the bermuda grass, Table 7. Both azithromycin and clarithromycin had been previously 398

detected in Hyperion biosolids (20). 399

400

The final analysis of data from Phase I, II, and III has shown the possibility, 401

although small, of transfer of specific ECs into select crops. The amount of ECs that 402

were transferred was minimal, part-per-trillion levels, but the likelihood does exist. 403

404

Although this study was designed to look at the possibility of transfer of human-405

use pharmaceuticals and other ECs into crops, the possibility exists for other avenues of 406

crop contamination via animal husbandry practices. Animal manures and composts are 407

widely used on both feed and food crops in irrigated desert production systems to 408

increase organic matter and improve overall soil fertility and tilth. Due to concerns of 409

microbial food risks, state programs such as the Arizona and California Leafy Greens 410

Marketing Agreements prohibit the application of raw manures for a one-year period 411

preceding the production of leafy vegetables. However, the programs do allow for 412

composted manure applications immediately before production, provided that testing 413

shows the food systems are free of coliform indicators. Composts are widely used by 414

organic producers as the principal forms of N and P fertilizers, and are also widely used 415

by conventional growers due to soil quality improvements and production benefits. 416

Therefore, further work with other pharmaceutical contaminants potentially present in 417

irrigation waters and animal husbandry waste composts (i.e., combined animal feed lots), 418

is warranted. 419

420

NOTICE: The United States Environmental Protection Agency through its Office of 421

Research and Development funded and managed the research described here. It has been 422

subjected to Agency’s administrative review and approved for publication. Mention of 423

trade names or commercial products does not constitute endorsement or recommendation 424

for use. 425

References

1. Wu, L.; Chen, W.; French, C.; Chang, A. Safe application of reclaimed water reuse in the

southwestern United States. ; 2009.

2. Bryck, J.; Prasad, R.; Davis, T. L. S.; Carpenter, G. National Database of Water Reuse

Facilities Summary Report; WaterReuse Foundation: Alexandria, VA, 2008.

3. Rowe, G. L.; Reutter, D. C.; Runkle, D. L.; Hambrook, J. A.; Janosy, S. D.; Hwang, L. H. Water

Quality in the Great and Little Miami River Basins, Ohio and Indiana, 1999-2001;

Circular 1229; Reston, VA, 2004.

4. Josephson, J., The Microbial "Resistome". Environmental Science & Technology 2006,

40, (21), 6531-6534.

5. Kümmerer, K., Resistance in the environment. Journal of Antimicrobial Chemotherapy

2004, 54, (2), 311-320.

6. Schwartz, T.; Kohnen, W.; Jansen, B.; Obst, U., Detection of antibiotic-resistant bacteria

and their resistance genes in wastewater, surface water, and drinking water biofilms.

FEMS Microbiology Ecology 2003, 43, (3), 325-335.

7. Schwartz, T.; Volkmann, H.; Kirchen, S.; Kohnen, W.; Schon-Holz, K.; Jansen, B.; Obst,

U., Real-time PCR detection of Pseudomonas aeruginosa in clinical and municipal

wastewater and genotyping of the ciprofloxacin-resistant isolates. FEMS Microbiology

Ecology 2006, 57, (1), 158-167.

8. Zhang, Y.; Marrs, C. F.; Simon, C.; Xi, C., Wastewater treatment contributes to selective

increase of antibiotic resistance among Acinetobacter spp. Science of the Total

Environment 2009, 407, 3702-3706.

9. Gilchrist, M. J.; Greko, C.; Wallinga, D. B.; Beran, G. W.; Riley, D. G.; Thorne, P. S.,

The potential role of concentrated animal feeding operations in infectious disease

epidemics and antibiotic resistance. Environmental Health Perspectives 2007, 115, (2),

313-316.

10. Batt, A. L.; Snow, D. D.; Aga, D. S., Occurrence of sulfonamide antimicrobials in private

water wells in Washington County, Idaho, USA. Chemosphere 2006, 64, (11), 1963-

1971.

11. Berger, K.; Petersen, B.; Buening-Pfaue, H., Persistence of drugs occurring in liquid

manure in the food chain. Archives fuer Lebensmittelhygeine 1986, 37, (4), 99-102.

12. Esiobu, N.; Armenta, L.; Ike, J., Antibiotic resistance in soil and water environments.

International Journal of Environmental Health Research 2002, 12, (2), 133-144.

13. Vaclavik, E.; Halling-Sorensen, B.; Ingerslev, F., Evaluation of manometric respiration

tests to assess the effects of veterinary antibiotics in soil. Chemosphere 2004, 56, (7),

667-676.

14. Baquero, F.; Martínez, J.-L.; Cantón, R., Antibiotics and antibiotic resistance in water

environments. Current Opinion in Biotechnology 2008, 19, (3), 260-265.

15. Andersen, S. R.; Sandaa, R. A., Distribution of Tetracycline Resistance Determinants

among Gram-Negative Bacteria Isolated from Polluted and Unpolluted Marine

Sediments. Applied and Environmental Microbiology 1994, 60, (3), 908-912.

16. Samuelsen, O.; Torsvik, V.; Ervik, A., Long-range changes in oxytetracycline

concentration and bacterial resistance towards oxytetracycline in a fish farm sediment

after medication. Science of the Total Environment 1992, 114, 25-36.

17. McKeon, D. M.; Calabrese, J. P.; Bissonnette, G. K., Antibiotic resistant gram-negative

bacteria in rural groundwater supplies. Water Research 1995, 29, (8), 1902-1908.

18. Loganathan, B.; Phillips, M.; Mowery, H.; Jones-Lepp, T. L., Contamination profiles and

mass loadings of select macrolide antibiotics and illicit drugs from a small urban

wastewater treatment plant. Chemosphere 2009, 75, (1), 70-77.

19. Jones-Lepp, T. L.; Alvarez, D. A.; Petty, J. D.; Huckins, J. N., Polar Organic Chemical

Integrative Sampling and Liquid Chromatography–Electrospray/Ion-Trap Mass

Spectrometry for Assessing Selected Prescription and Illicit Drugs in Treated Sewage

Effluents. Archives of Environmental Contamination and Toxicology 2004, 47, (4), 427-

439.

20. Jones-Lepp, T. L.; Stevens, R., Pharmaceuticals and personal care products in

biosolids/sewage sludge: the interface between analytical chemistry and regulation.

Analytical and Bioanalytical Chemistry 2007, 387, (4), 1173-83.

21. USEPA Targeted National Sewage Sludge Survey Report.

http://earth1.epa.gov/waterscience/biosolids/tnsss-fs.html,

http://www.epa.gov/waterscience/biosolids/tnsss-overview.html,

http://www.epa.gov/waterscience/biosolids/

22. Boxall, A. B.; Johnson, P.; Smith, E. J.; Sinclair, C. J.; Stutt, E.; Levy, L. S., Uptake of

veterinary medicines from soils into plants. Journal of Agricultural and Food Chemistry

2006, 54, (6), 2288-97.

23. Migliore, L.; Cozzolino, S.; Fiori, M., Phytotoxicity to and uptake of enrofloxacin in crop

plants. Chemosphere 2003, 52, (7), 1233-1244.

24. Grote, M.; Schwake-Anduschus, C.; Michel, R.; Stevens, H.; Heyser, W.; Langenkamper,

G.; Betsche, T.; Freitag, M., Incorporation of veterinary antibiotics into crops from

manured soil. Landbauforschung Volkenrode 2007, 57, (1), 25-32.

25. Dolliver, H.; Kumar, K.; Gupta, S., Sulfamethazine Uptake by Plants from Manure-

Amended Soil. Journal of Environmental Quality 2007, 36, (4), 1224-1230.

26. Herklotz, P. A.; Gurung, P.; Vanden Heuvel, B.; Kinney, C. A., Uptake of human

pharmaceuticals by plants grown under hydroponic conditions. Chemosphere 2010, In

Press.

27. Sanchez, C. A.; Barraj, L. M.; Blount, B. C.; Scrafford, C. G.; Valentin-Blasini, L.;

Smith, K. M.; Krieger, R. I., Perchlorate exposure from food crops produced in the lower

Colorado River region. Journal of Exposure Science and Environmental Epidemiology

2008, 19, 359-368.

28. Sanchez, C. A.; Blount, B. C.; Valentin-Blasini, L.; Lesch, S. M.; Krieger, R. I.,

Perchlorate in the Feed Dairy Continuum of the Southwestern United States. Journal

Agricultural and Food Chemistry 2008, 56, 5443-5450.

29. Sanchez, C. A.; Jones-Lepp, T.; Wilson, D.; Alvarez, D. In Pharmaceuticals in Waste

Streams and Surface Waters of the Colorado River Basin, Lake Mead Science

Symposium, Las Vegas, NV, 2009; Las Vegas, NV, 2009.

Tables

1. Chart of samples collected.

2. Tucson wastewater effluent ECs concentrations.

3. Chemical and physical properties of loam soil used in field experiment Phase II (UA-CAC)

and Phase III (UA-YAC).

4. Emerging contaminants, MW, precursor and product ions, and LODs.

5. Accuracy and precision spiked recovery parameters (0.5 μg g-1

and 1 μg g-1

) from Bermuda

grass, lettuce, spinach, carrots (roots and leaf).

6. Phase I – Results from greenhouse study.

7. Phase II and III – Results from UA-CAC field study and UA-YAC field study.

Table 1. Chart of samples collected.

Sample type Phase I

Greenhouse

Phase II

UA-CAC

Phase III

UA-YAC

Bell pepper, green X

Bermuda grass X

Cantaloupe X

Carrots X X

Lettuce X

Spinach X X X

Soils X X X

Water

IDD

Tucson WWTP

Tucson well water

X

X

X

X

Watermelon X

UA-CAC = University of Arizona Campus Agricultural Center, Tucson, AZ

UA-YAC = University of Arizona Yuma Agricultural Center

IDD = Imperial Diversion Dam, Colorado River

Table 2. Tucson wastewater effluent ECs concentrations.

collection

date

ng L-1

azithromycin roxithromycin clarithromycin methamphetamine n,n'-DMPEA pseudoephedrine

Tucson effluent 02/28/08 255 ND ND 144 ND 566

Tucson dup 02/28/08 255 ND ND 222 ND 713

Tucson well 03/24/09 ND ND ND ND ND ND

Tucson effluent 04/01/08 686 ND ND 288 ND 680

Tucson effluent 04/10/08 162 880 ND 155 21 229

Tucson effluent 04/29/08 323 ND ND 99 ND 86

Tucson effluent dup 04/29/08 285 ND ND 135 ND 76

Tucson effluent 05/29/08 259 ND ND 309 ND 158

Tucson effluent dup 05/29/08 267 ND ND 289 ND 216

Tucson well 05/29/08 ND ND ND ND ND ND

Tucson effluent 07/02/08 176 ND ND 568 ND 608 ND = not detected. n,n’-DMPEA = n,n’-dimethylphenethylamine

Table 3. Chemical and physical properties of loam soil used in field experiment Phase II & Phase III.

Parameter (unit) Value

Phase II Phase III

pH 8.2 7.8

EC (dS m-1

) 2.2 1.8

ESP (%) 3.3 4.1

Organic C (%) 0.7 1.2

Sand (%) 50 9

Silt (%) 38 53

Clay (%) 12 38

Nitrate-N (mg kg-1

) 42 20

Bicarbonate Soluble Phosphate (mg kg-1

) 24 27

EC=electrical conductivity and ESP=exchangeable sodium percentage.

Table 4. Emerging contaminants, molecular weight, precursor and product ions, and LODs. Analyte

CAS #

Molecular

weight (amu)

Precursor ions Product ion

(confirmation ions)

LOD

ng,

on-column

Azithromycin

(83905-01-5)

748.5 749.5 (M+H)+ 591.4 (M+H-C8H16O2N)

+ 0.5

Roxithromycin

(80214-83-1)

836.5 859.5 (M+Na)+ 755.4 (M+Na-C4H9O3)

+ 1

Clarithromycin

(81103-11-9)

747.5 748.4 (M+H)+ 590.1 (M+H-C8H16O2N)

+ 1

Clindamycin

(18323-44-9)

424.2 425.2 (M+H)+ 377.2 (M+H-SH-CH3)

+ 1

Methamphetamine

(537-46-2)

149.3 150 (M+H)+ 119 (M+H-CH3NH2)

+ 1.5

MDMA

(69610-10-2)

193 194 (M+H)+ 163.0 (M-CH3NH2+H)

+ 1

Pseudoephedrine

(90-82-4)

165.2 166 (M+H)+ 148.2 (M+H-H2O)

+ 0.5

n,n-dimethylphenethylamine

(1126-71-2)

149.2 150 (M+H)+ 105 (M-N(CH3)2)

+ 0.5

MW = molecular weight; LOD = limit-of-detection

Table 5. Accuracy and precision spiked recovery parameters (0.5 μg g-1

and 1 μg g-1

) from Bermuda grass, lettuce, spinach, carrots.

% Recovery (standard deviation; % relative standard deviation) †

Sample type

Compound

Bermuda

Roots

Lettuce leaf Lettuce

root Spinach leaf Spinach root Carrot root Carrot tops

Azithromycin 20 (± 4; 20%) 22 (± 2; 10%) 2 (± 1) 45 (± 9; 20%) 5 (± 1; 20%) 19 (± 6; 32%) 19 (± 1; 5%)

Roxithromycin 40 (± 3; 8%) 32 (± 5; 16%) 26 (± 2) 29 (± 4; 14%) 48 (± 4; 8%) 76 (± 17; 23%) 35 (± 5; 13%)

Clarithromycin 22 (± 6; 25%) 20 (± 2; 11%) 10 (± 1) 22 (± 4; 20%) 16 (± 3; 17%) 32 (± 9; 28%) 21 (± 3; 12%)

Clindamycin 33 (± 7; 22%) 30 (± 8; 26%) 22 (± 1) 23 (± 6; 26%) 38 (± 9; 24%) 35 (± 5; 15%) 32 (± 4; 12%)

Methamphetamine 44 (± 6; 14%) 24 (± 4; 16%) 15 (± 0) 21 (± 2; 7%) 33 (± 9; 28%) 30 (± 4; 15%) 36 (± 5; 13%)

MDMA 45 (± 8; 17%) 23 (± 1; 6%) 11(± 0) 23 (± 4; 18%) 22 (± 15; 69%) 26 (± 6; 21%) 26 (± 1; 4%)

n,n'-dimethylphenethylamine 47 (± 10; 21%) 29 (± 1; 5%) 17 (± 2) 22 (± 3; 13%) 23 (± 6; 28%) 29 (± 5; 16%) 38 (± 1; 2%)

Pseudoephedrine 50 (± 3; 6%) 27 (± 0; 0%) 17 (± 1) 24 (± 2; 8%) 20 (± 15; 74%) 23 (± 6; 28%) 28 (± 1; 4%) †n=3 for all sample types, except lettuce roots: n= 2; and carrot roots: n=6

Table 6. Phase I – Results from greenhouse study.

ng g-1

, n = 2

Spiked

Compound

Lettuce

leaf

Lettuce

Root

Spinach

leaf

Spinach

root

Carrot

greens

Carrot

root

Azithromycin ND ND ND ND ND ND

Roxithromycin ND < 10 ng g-1

LOQ ND ND ND 115

Clindamycin ND < 10 ng g-1

LOQ ND < 10 ng g-1

LOQ* ND 53

ND = not detected; *not enough sample for duplicate extraction.

Table 7. Phase II and III – Results from UA-CAC* field study and UA-YAC** field study. ng g

-1

Sample type

Analyte

Detected

Bermuda

Grass**

Bermuda

roots**

Cantaloupe*

Carrot

Roots*

Green bell

pepper*

Spinach*

Spinach**

Watermelon*

n,n'-DMPEA ND 125 53 ND 58 48 ND 180

Azithromycin ND 90 ND ND ND ND ND ND

Clarithromycin ND 135 ND ND ND ND ND ND ND = not detected; n,n’-DMPEA = n,n’-dimethylphenethylamine